Laserfiche WebLink
<br />B-8 <br /> <br />* Israeli II * <br /> <br />The use of a cross-over design in the earlier Israeli I led to some concern about remote <br />effects between one pair of target and control areas and the other. Presently, there seems to be <br />a reasonable consensus among meteorologists that serious remote effects between the Israeli I <br />area pairs are quite unlikely, although some analyses have indicated that there may have been <br />some contamination of the North target during some Center seeding days. This would have <br />reduced sensitivity, but should not affect validity of a positive result. <br /> <br />In Israeli II, an area in the South was seeded on those days when the main target area was <br />not seeded. Thus remote-effect issues could arise. To the extent to which the meteorological <br />judgments about Israeli I also apply to the altered location of Israeli II, these issues may not be <br />serious and may, perhaps, even be neglected. <br />... Alberta hail * <br /> <br />Remote effects are either not a problem, or submerged by such problems as lack of ran- <br />domization. <br /> <br />... Santa Barbara II * <br /> <br />Control area for phase 1 was upwind enough to avoid microphysical effects. Since it was <br />not used in the analyses to which we give most weight, remote-effect issues seem unimportant. <br /> <br />* South Africa hail * <br /> <br />No control areas, and hence no remote effect issues (except for possible effects from <br />other seeding operations). <br /> <br />* North Dakota * <br /> <br />No control areas, and hence no remote-effect issues (except for possible effects from <br />other seeding operations). <br /> <br />* Tasmania ... <br /> <br />This study depended very heavily on control areas to provide precision of results. Seeding <br />was interrupted -- or aborted -- whenever wind directions suggested that AgI nuclei might be <br />conveyed directly toward a control area. (Since such days were still counted as "seeded days", <br />no false significances would result.) In view of the limited area offered by the interior of this <br />large island, remote-effect issues must, we feel, be taken seriously. We do not feel competent <br />to adequately evaluate them here. <br /> <br />11. Statistical issues <br /> <br />... FACE * <br /> <br />This study has so far been exploratory, with the major purely statistical questions revolv- <br />ing about the large contribution of a few days of apparent high success. If this concentration of <br />apparent effect on a few days continues in the proposed confirmatory phase, that phase may <br />have to be extended longer than otherwise would have been required. <br />* NHRE ... <br /> <br />Rainfall results up to the point of early termination are not close to significance. (Rainfall <br />analyses were added after the fact.) <br /> <br />, <br /> <br />* Colorado River Basin * <br />